Over voltage report - is it real?

[Brough]

We have a WS-6-Mini that's powered over a 3 ft cable by a Ubiquiti AF24 PoE injector, i.e. 50v and up to 1.2A. It in turn powers a Ubiquiti PBE-5AC radio and a MikroTik RB2011 router. The site was installed 21 months ago and has operated w/o problems until a week ago when we got over-voltage alarms from the mini-switch. The site is still on-line and behaving normally except for the over-voltage indication and an unusually high wattage reported on port 3. I suspect the excess wattage on port 3 is due to fried current sensors. I don't know whether to believe the over-voltage readings or not. If true, at 30.1v the 24v supply is way over-voltage.

The MT router specs say it's good to 30v. But I'm under the impression that >28v would fry a Ubiquiti radio. Unfortunately, ubnt doesn't provide useful specs so I'm relying on on-line discussions.

Since the site is still working in all other regards, it may be a few more days before someone goes to investigate. Is it possible the voltages are ok and only the measurements have failed?

[sirhc]

Your unit is damaged, most commonly caused by ground current which fries a current or voltage sensor. However since all sensors share the same bus if one gets damaged in the right way it affects others.

When you swap it out of service you can open it up and examine the board (both sides) and the damage should be visible.

Keep in mind GROUND CURRENT is AC voltage normally 110V but often the frequency is much higher than 50/60Hz which allows the current to pass through filter CAPs designed to squelch noise on the line. This AC current does not have to be high amperage to cause a lot of damage. Lab tests show current sensors blown off the board with as little as .1 milliamp. Keep in mind your talking about 120V AC flowing through the board > 50/60M Hz sometimes as high as 500 Hz or higher from harmonics caused by failing electric motors compressors or the current comes in from the service provider as the electric company hands off (2) hot, 1 neutral, and 1 ground and all grounds in an area are bonded together.

Now if your electrical service rods are insufficient to absorb the current and the tower rods are seen as better path to ground the current traverses the switch/cable to get to tower grounding system. Bonding all ground services helps but not 100%. Anytime you lose equipment from suspected ground current add 2 more service ground rods. I have a tower on a mountain top with rocky soil providing poor ground, took 6 additional ground rods before problem stopped.

Ground current damage often occurs during rain events as the rain envelopes the ground rods reducing their capacity or increasing their resistance to ground.
READ THIS POST:
https://forum.netonix.com/viewtopic.php?f=17&t=5495&p=29602&hilit=+ground+current+Distributor#p29602

[waveband1]

Hi All, I am seemingly having the same issues, I believe, and was curious to see if I can find our more information. We're new the Netonix, this is our second Mini in use on a building roof top, the first switch is performing flawlessly but this one seems to be having some heartburn.

Considerations: The Mini is being powered on Port 1 and is simply powering the Mini, and two (2) Mikrotik 60G (RBwAPG-60) and off the switch we service three customers. The installer ran all the cables 'shielded' ( not sure if this is affecting anything).


Peculiarities -
> The three (3) shielded customer cables (ports 4, 5, 6) , are displayed in Yellow and show as 'bad' on cable diagnostics with different length segments, but they test fine on the Fluke.
> Power> Board 48V is 100% red reading "20.4"
> Received log entries:
>> with error codes: switch[766] 48V is out of range (20.4v)
>> with error codes: switch[766] 24V is out of range (20.5v)

Any thoughts as to whether the device is corrupt or if it's the power source, which I believe to be 24V, grounding maybe? anything would be great, sorry - I was going to post some pics but I can seem to figure that out.

Thanks!

[sirhc]

Hi All, I am seemingly having the same issues, I believe, and was curious to see if I can find our more information. We're new the Netonix, this is our second Mini in use on a building roof top, the first switch is performing flawlessly but this one seems to be having some heartburn.

Considerations: The Mini is being powered on Port 1 and is simply powering the Mini, and two (2) Mikrotik 60G (RBwAPG-60) and off the switch we service three customers. The installer ran all the cables 'shielded' ( not sure if this is affecting anything).


Peculiarities -
> The three (3) shielded customer cables (ports 4, 5, 6) , are displayed in Yellow and show as 'bad' on cable diagnostics with different length segments, but they test fine on the Fluke.
> Power> Board 48V is 100% red reading "20.4"
> Received log entries:
>> with error codes: switch[766] 48V is out of range (20.4v)
>> with error codes: switch[766] 24V is out of range (20.5v)

Any thoughts as to whether the device is corrupt or if it's the power source, which I believe to be 24V, grounding maybe? anything would be great, sorry - I was going to post some pics but I can seem to figure that out.

Thanks!

Your unit is damaged, most commonly caused by ground current which fries a current or voltage sensor. However since all sensors share the same bus if one gets damaged in the right way it affects others.

When you swap it out of service you can open it up and examine the board (both sides) and the damage should be visible.

Keep in mind GROUND CURRENT is AC voltage normally 110V but often the frequency is much higher than 50/60Hz which allows the current to pass through filter CAPs designed to squelch noise on the line. This AC current does not have to be high amperage to cause a lot of damage. Lab tests show current sensors blown off the board with as little as .1 milliamp. Keep in mind your talking about 120V AC flowing through the board > 50/60M Hz sometimes as high as 500 Hz or higher from harmonics caused by failing electric motors compressors or the current comes in from the service provider as the electric company hands off (2) hot, 1 neutral, and 1 ground and all grounds in an area are bonded together.

Now if your electrical service rods are insufficient to absorb the current and the tower rods are seen as better path to ground the current traverses the switch/cable to get to tower grounding system. Bonding all ground services helps but not 100%. Anytime you lose equipment from suspected ground current add 2 more service ground rods. I have a tower on a mountain top with rocky soil providing poor ground, took 6 additional ground rods before problem stopped.

Ground current damage often occurs during rain events as the rain envelopes the ground rods reducing their capacity or increasing their resistance to ground.
READ THIS POST:
https://forum.netonix.com/viewtopic.php?f=17&t=5495&p=29602&hilit=+ground+current+Distributor#p29602

If you have a unit that is not working right RMA the unit. We will provide a full report of what we find. If it failed we will repair it free and return, if it is damaged we provide a full report with pictures of damage and our best guess as to the cause and there is a fee for the repair.

Step 1 is to RMA the unit.
Even if a unit is out of warranty period we will still repair it for a fee which is cheaper most times than buying anew unit.

[jakematic]

> Power> Board 48V is 100% red reading "20.4"
> Received log entries:
>> with error codes: switch[766] 48V is out of range (20.4v)
>> with error codes: switch[766] 24V is out of range (20.5v)

Any thoughts as to whether the device is corrupt or if it's the power source, which I believe to be 24V, grounding maybe?

WS-6-Mini needs 48V or 48VH.

If you are powering it with a 24V brick that could explain the voltages.


https://www.netonix.com/wisp-switch/ws-6-mini.html

[sirhc]

Jake you are partially correct, and you caught the possibility he is trying to power with 24V.

Yes needs 48V - 54V to power up.

But please do not power the switch with 48V single pair power brick or POE option from NEtonix switch.

MUST BE 48VH and or POE Brick we sell or UBNT airFIBER 24 POE brick which powers on all 4 pair or you will damage switch.

[waveband1]

Thanks everyone, interestingly enough I did start with the Netnoix purchased NPOE-54V-80W; however it did not power the switch, I received some sort of error; if I recall it was a distance error?, I think the run may be ~150-200 ft... not sure should've wrote it down - but then I switched it to, you're correct, a 24V brick since I figured it was only powering a couple of small Tik AP's and it worked... well seemingly anyway...

Currently it's working and stable, so not sure if I should just rip and replace or try to swap out the power but thanks for the input!

[jakematic]

Currently it's working and stable, so not sure if I should just rip and replace or try to swap out the power but thanks for the input!

If it shows 100% on 48VDC then it is not stable, and 24V is not recommended.
Got one about 150' away from me right now powered from a Netonix brick and it works perfectly powering a mix of 48V and 24V devices.
Two others are powered 48VH from WS-12 and also are very stable - never see a blip in the logs.

Suggest powering it otherwise, and/or RMA if your grounds aren't correct as Chris said.

[jorgenekah]

Hi, I haave yesterday installed a brand new netonix form ISP supplies. It is a WS-12-250-DC . I am using rocket radios which normally consumes 3 to 7watts, but on port 8 it shows 40watts and 27watts on port 9 which is not ok, again it is a brand new switch.
Please help me to fix this what can I do, firmware is 1.5.5

[Dave]

sorry, it sounds like that switch has been damaged somehow, you will need to RMA it so we can determine what happened.